Limitations included in this standard for various prequalified connections include specification of permissible materials for base metals, mechanical properties for weld filler metals, member shape and profile, connection geometry, detailing, and workmanship. These limitations are based on conditions demonstrated by testing and analytical evaluation for which reliable connection behavior can be attained. It is possible that these connections can provide reliable behavior outside these limitations; however, this has not been demonstrated. When any condition of base metal, mechanical properties, weld filler metals, member shape and profile, connection geometry, detailing, or workmanship falls outside the limitations specified herein, project-specific qualification testing should be performed to demonstrate the acceptability of connection behavior under these conditions.
Limited testing of connections of I-shaped beams to the webs of I-shaped columns had been conducted prior to the Northridge earthquake by Tsai and Popov (1986, 1988) and others. This testing demonstrated that these “minor-axis” connections were generally incapable of developing reliable inelastic behavior. Only very limited testing of such minor-axis connections has been conducted since the Northridge earthquake. Consequently, the majority of connections in this standard are not prequalified for I-shaped beams framing into the web of an I-shaped column. The exceptions to this are the SidePlate moment connection and the DuraFuse Frames moment connection. Both of these connections have prequalified configurations for connecting an I-shaped beam to the web of an I-shaped column.
Although there has been only limited testing of connections in assemblies subjected to biaxial bending of the column, the judgment of the CPRP was that as long as columns are designed in accordance with the AISCE Seismic Provisions, which allows only limited yielding in columns, it would be possible to obtain acceptable behavior of beam-column connection assemblies subjected to biaxial loading.
Nearly all moment frame connection tests have been performed on single- or double-sided beam-column subassemblies with the beam perpendicular to the vertical axis of the column (i.e., a level beam) and coplanar with the strong axis of the column (i.e., unskewed). The reality of building construction is that sloping beams occur in most structures, such as at the roof. Occasionally, skewed configurations are required to accommodate architectural requirements. This standard does not contain provisions that explicitly address sloped or skewed moment frame beams because of the wide variety of connections that have obtained prequalification and the lack of systematic physical testing for each connection. Professional judgment, therefore, is required to determine whether a proposed frame geometry is appropriately covered by the prequalification limits in this standard.
One factor to consider when evaluating the impact of frame beam slope is the absolute angle at which the beam slopes. For example, even so-called “flat” roofs slope at approximately in./ft (21 mm/m), which equates to an angle of . Testing of frame beams using reduced beam sections sloping at (Ball et al., 2010) indicates that at this angle, the performance of the connection can be adversely affected depending on how it is configured. The former angle likely is not significant while the latter angle appears to be, at least for the reduced beam section. It is reasonable to assume that different connections will likely have different thresholds above which the slope becomes significant because of characteristics that govern connection behavior.
For example, connection geometry and limit states that govern connection behavior will likely influence the threshold above which frame beam slope becomes significant. Connection performance of sloped beams is expected to be influenced by components of differing length or changes in geometry (e.g., the relative distance from the face of column to the first bolt or the length of the flange plate). These variations may impose different levels of load on components at the top flange versus those on the bottom flange due to changes in relative stiffness. The relative onset of limit states, such as flange local buckling, prying, and similar may be influenced by the angle of slope and is expected to vary from connection type-to-connection type.
Limited analytical studies have been performed relative to the impact of beam skew on connection performance. Prinz and Richards (2016) report that frame beam skew angles between 10° and 20° in reduced beam section connections appear to cause limited increases in column torsional demand and limited additional flange tip yielding. They also report that skew angles of 10° reduce low cycle fatigue capacity in the reduced section region by less than one cycle. Thus, skew angles of less than 5° to 10° might be considered acceptable in reduced beam sections connections. Similar analytical studies have not been conducted for other connection geometries.